For a pdf of the report with charts see "Deeper Insights on the Australian Labour Market" (pdf 304kb)

March employment beat expectations taking the level back above where it was pre-COVID with total hours worked surging higher as well; the participation rate hit a historic high, due to a record high in female participation; while the unemployment rate fell to 5.6%, still 0.4ppt higher than last March.
We were surprised by the strength in some of the indicators; in particular, the surge in hours worked and the fall in zero hours worked back to pre-COVID levels.
We used those indicators to guide our estimate of the impact of the ending of JobKeeper. Such positive prints suggest the impact could be far smaller than our -100k estimate.
Female participation hit a record high in March but it is the stickiness in male participation, with a poor recovery in male full-time employment, that is behind an unemployment rate 0.4% higher than a year earlier. Despite a record high in participation, female unemployment is lower than male unemployment.
The outperformance in female employment (both full-time and part-time) is not explained by growth in industries that have a higher share of female employment but rather a broader underperformance in male employment (in particular full-time employment).
The tightening in the Victorian labour market should set the scene for a lift in wage inflation in that state later this year but even there, at best a pace of 2%yr heading into 2021 is the likely outcome.
As the impact of the COVID lockdowns first appeared in the April 2020 Labour Force Survey, through-the-year comparisons of the various labour market indicators can provide a picture of what is driving, and what is lagging, the current recovery.

Total employment rose 70.7k in March, firmly above our estimate of 32.0k and the market median of 35.0k (the upper end of the market range was 65.0k).

This strong jobs gain was enough to push the unemployment rate down to 5.6%, despite participation rising 0.2ppts to a record 66.3%, driving a 43.6k lift in the labour force.

Job gains were concentrated exclusively in part-time employment, which advanced 91.5k and more than offset a 20.8k fall in full-time. This drove part-time employment up almost 77k (+1.9%yr) compared to a year ago, while full time is 2.5k (0.0%) below March 2020.

While some of the fall in full-time, and rise in part-time, would be associated with employees moving between full-time and part-time classification as their hours worked changed this does not change the overall picture that the recovery in part-time employment continues to outpace the recovery in full-time employment.

Those that were employed, but worked zero hours for economic reasons, fell from 107.0k to 61.2k. Zero hours employees totalled 79.5k in January, peaked at 731.5k in April 2020 and averaged 55k per month in the year before Covid. This data has been seasonally adjusted by Westpac.

The fall in the unemployment rate was accompanied by a significant reduction in underemployment, which was down an outsized 0.6ppts to 7.9%. This leaves underemployment 1.1% below its pre-Covid (March 2020) level. Underutilisation also came off 0.8ppts to 13.5%.

In March, hours worked advanced a further 2.2%, compounding the +6.1% reversal in February as employees returned from the holiday period. Hours worked are now above pre-Covid levels, up 1.2% compared to March 2020.

Results were mixed across the states. NSW realised an employment gain of 0.4%, which saw the unemployment rate drop 0.3ppts to 5.4%. Victoria, conversely, saw unemployment rise 0.4%, which may be reflecting ongoing impacts of a circuit-breaker lockdown late in February. Qld saw unemployment down 0.3%, with reopening borders and the tourism package possibly driving some labour-hoarding. SA and WA – two states that have seen very low population growth throughout the pandemic (which is a helpful dynamic for the unemployment rate) – saw significant reductions in unemployment, down 0.5ppts and 1.3ppts respectively.

Recovery in female employment has been stronger and more balanced compared to male employment

In March, we continued to see female unemployment falling, down 0.2ppts to 5.2%. This was despite a 0.4ppt increase in female participation. Male unemployment also fell 0.2ppts to 5.9%, but participation was flat over the month.

Female employment lifted 74.3k/0.6% in the year to March and while this is split between full-time and part-time, the recovery favoured full-time with a +44.9k/1.3% gain in the year compared to a 30.1k/1.1% rise in part-time employment. Male employment was flat in the year to March (-0.7k) with a -0.9%/-47.4k loss in full-time employment offset by a 3.6%/46.8k gain in part-time employment.

One could argue that this may be due to labour hoarding with firms reducing hours worked for many employees as they try to hold on to workers and prepare for the ongoing recovery. If this was so, you would expect to see a rise in male underemployment as those now working part-time would be willing and able to work more hours if offered. However, male underemployment has fallen 0.6ppt to 6.6% in the year to March suggesting a greater share of males are finding the work they desire despite the rising share of part-time employment. Contrast this with female where the stronger recovery in full-time employment may be why there was a larger 1.3ppt fall in underemployment to 9.3% in the year to March. We also note that the higher level in female underemployment is historical and likely reflects the higher share of female part-time employment.

Female unemployment lower than males

During the COVID recession participation has been quite dynamic. Historically, female workers enter and leave the workforce with shifts in labour demand in a more dynamic way than males but the greater than usual magnitude of the shifts in employment through 2020 was also reflected in greater-than-usual shifts in participation for both genders.

The rise in participation in the year to March has been stronger for females compared to males, 0.6ppt to 61.8% (a record high) vs. 0.2ppt to 70.9%. It was this strong lift in female participation that has moderated the improvement in the female unemployment rate with the female labour force lifting 1.4%/92.4k in the year to March compared to a 1.2%/74.9k gain in employment. As a result the female unemployment rate at 5.3% is still 0.2% higher that it was a year earlier.

While smaller than the gain in female participation, male participation lifted 0.2ppt in the year to March which was enough to add 0.6%/44.0k to the male labour force outpacing the flat print on male employment. As such, male unemployment lifted 0.6ppt in the year to March and at 5.9%, it is higher than female unemployment.

So while female participation hit a record high boosting female unemployment it appears that the stickiness in male participation, in the light of poor full-time employment growth, is the more significant factor for why total unemployment is meaningfully higher that it was last March.

Strength in female employment is due to a broader underperformance in male employment

It is true that most of the industries that made the greatest gains in the year February (industry data is only up to February in Figure 8) are sectors that have a high share of female employment; retail trade has added 70.2k in the year and the female employment share is 54%. However, in the year retail added 63k male employees and only 7.7k female employees. Professional & scientific also has a relatively high share of female employment at 42% at it lifted employment 58k in the year but it too has seen a larger gain in male employment in the year (+46k) than in female employment (+12k). It was public administration (50% female) that made a sound 29k contribution to employment growth in the year and all of this was female (+39.6k) offsetting the fall in male employment (-10k).

Looking at the sectors that are still down in the year, accommodation & food services (-83k) employment is 45% female and there female employment is still down -55k/-11% while male employment is down -28k/-7%; administration & support services employment is 50% female and the -55k loss of employment in the year is split fairly evenly between the genders while education & training (72% female) is still down 37k employees by males have been hit quite a bit harder (-29k) than female (-9k).

It does not appear that you can use the argument that the stronger gains in female employment is compositional, that is it is due to employment growth in sectors that have a higher share of female employment, but rather a broader underperformance of the recovery in male employment. We expect to see some improvement in male employment as construction, manufacturing and distribution employment picks up through the year but there is nothing to suggest the broader outperformance in female employment is likely to fade anytime soon.

Employment growth diverges between states

The employment recovery in NSW and Victoria has lagged the recovery seen across the rest of the nation, suggesting there may be more scope for catch-up across these two states. In NSW and Vic., higher caseloads, longer lockdowns, and more reliance on international travel buffeted the recovery over the second half of 2020, which constrained state demand and jobs growth. Conversely, the smaller states had a milder experience with the virus and a shorter period under lockdown, whilst still benefiting from record levels of fiscal and monetary stimulus.

As seen in Figure 9 and 10, the two-speed labour market recovery (and the scope for "catch up") closely maps the profile of state demand. Between Q4 2019 and Q4 2020, activity in NSW and Victoria together contracted by 1.9%, but across the rest of the nation, output expanded by 1.9%. Similarly, employment growth between December 2019 and March 2020 in NSW and Victoria was essentially flat, but rose 1.9% across the other states.

The recovery in wages is likely to be led by Victoria

The dispersion in relative state growth and labour market performance matters particularly in regards to the outlook for wages. Victorian average hourly wage rates have been more responsive to changes in labour market conditions than have wages in NSW (Figure 11). As such, the recent improvement in labour market conditions, with the Victorian underutilisation rate falling back to close to where it was pre-COVID, is already pointing to wage inflation in that state lifting from the current record low of 1.1%yr back to around 2% by late 2021. By contrast, wage inflation in NSW is slightly stronger but the lower underutilisation rate in that state is also consistent with around 2% pace in wages at end 2021. A lot will depend on what happens with the minimum wage increase this year, our assumption is that it returns to an around 3% increase, and how much pressure the lack of supply of international labour will have on wages negotiable by individual arrangements. Naturally, a smaller than expected increase in the minimum wage will hold back wage inflation while there is clear risk the lack of supply in labour in particular industries/skills sets/regions could see a boost to wage inflation by more than anticipated, especially while the international borders remain closed.

Indicators point to smaller impact from end of JobKeeper

In our Economic Bulletin published 29 March, we looked at several labour market indicators to estimate the impact of the conclusion of JobKeeper (JK). One of these measures, proposed by Prof. Jeff Borland (Labour Market Snapshot #76, "Another Step Forward But How Many Steps Backwards Are We About To Go?" March 3, 2021) infers labour market stress from the wedge between the growth in employment and hours worked since the beginning of the pandemic. The rationale is that if hours worked has fallen markedly, but employment has held up, the wedge between these variables should represent the supportive effect of JK in sustaining employment – that is, without JK, employment would have fallen more in line with the fall in hours worked.

As shown in Figure 12 the February labour force survey revealed that employment had recovered to pre-pandemic levels, whilst hours worked was down 0.67%. By multiplying this "wedge" by total employment, we inferred that JK was supporting about 90,000 jobs in February (which, by extension, could be vulnerable when the scheme ended).

In the March labour force survey, this wedge has closed: compared to pre-Covid levels, employment was up 0.6% while hours worked was up 1.2%. Whilst this does not mean that there will be no job losses due to the ending of JobKeeper (there will be pockets of stress in some sectors), the strong recovery in hours worked suggests that the labour market is in a much more robust position than we thought even just a month ago.

Justin Smirk, Senior Economist +61 (2) 8254 9336

Lochlan Halloway Economist +61 (2) 8254 0104

Deeper Insights on the Aust. Labour Market

Related articles

Padlock {{article.title | ellipsis}}

{{author}} & , {{article.authors[0]}}, {{article.authors[1]}}, {{article.authors[2]}}, ...{{author}} and ,

{{article.publishDate | friendlyDate}}

{{article.title | ellipsis}}

{{author}} & ,

{{article.authors[0]}}, {{article.authors[1]}}, {{article.authors[2]}}, ...{{author}} and ,